AT&T Integrated Speech Research Workstation

The tools currently available to many speech researchers are disjoint, requiring the user to work in a variety of hardware and software environments. In a typical scenario PCs with data acquisitions cards are used to collect speech. Graphics workstations are used to display and edit speech. Powerful computers are used to train the templates or models. Finally, specialized recognition hardware is used to execute algorithms. To use all of these different machines, the researcher is required to move data between them and to understand the different hardware and software environments.